Hissing and whistling in the STe audio output

[exxos]

Hold up, where is the 2600hz noise in those test then ?

Can you check pin 8 of the last op amp output as well ?

[sandord]

Hold up, where is the 2600hz noise in those test then ?

My thoughts exactly! Will do some more measuring today, including the cinches for comparison.

[exxos]

My thoughts exactly! Will do some more measuring today, including the cinches for comparison.

Also worth noting if the psu being physically close to the amps or not.... noise could come from the psu.. so would have to by physically close to where it normally sits...

[sandord]

Also worth noting if the psu being physically close to the amps or not.... noise could come from the psu.. so would have to by physically close to where it normally sits...

Yep, I was going to mention that (my psu was sitting outside the casing) but wanted to measure the difference first.

I will probably post more results this afternoon.

[sandord]

These are the points where I measured in the audio path (all with the PSU outside of the ST case):

ste2.png (204.97 KiB) Viewed 5322 times

I hope I got the locations right, here's a picture for you guys to check:

ste1.png (1.37 MiB) Viewed 5322 times

I've found a nice realtime FFT program (http://www.techmind.org/audio/specanaly.html).
The gray line is the average, the black line the current signal. I've made the screen shot with the lead disconnected so in this case, the gray signal is the real signal, the black one is de base line (wire noise).

ste3.png (34.15 KiB) Viewed 5322 times

ste4.png (32.85 KiB) Viewed 5322 times

ste5.png (32.01 KiB) Viewed 5322 times

ste6.png (31.62 KiB) Viewed 5322 times

Point D gives a rather interesting signal. I don't see the 2600Hz spike but the signal is suspicously 'clean'.

[exxos]

Very good info. What is that MF4 part ? Noise seems to start on its output then get amplified..

[Stimpy]

Is it an anti alias filter or something? Perhaps the clock driving it needs to be series terminated? Or maybe its just a noisy chip?

[sandord]

Is it an anti alias filter or something? Perhaps the clock driving it needs to be series terminated? Or maybe its just a noisy chip?

It's a 4th order switched capacitor Butterworth lowpass filter.

See https://pdf1.alldatasheet.com/datasheet ... N-100.html

[sandord]

Signal C is the filter out (pin 5) of the MF4.

I just measured filter in (pin 8) and I can see a little spike at 2600Hz as well there. I can't get a clean screen shot of this since the average line is hidden behind the current signal.

Now tracing further back into the audio path...

[exxos]

What is on FCLK ?